Satellites have tangible impact on our daily lives; they revolutionized our everyday living. Satellite batteries are expected to deliver the power demand at any time during the period of an eclipse or when the power received from the solar panel is not sufficient. This study is performed in order to develop a scheduling algorithm that augments the runtime/lifetime of the battery; which will consequently aim on diminishing the State of Health (SOH) degradation of the battery. Data will be collected for an existing satellite, such as Nayif-1, in order to analyze battery behavior in space. In parallel, an accurate State of Charge (SOC) estimation technique will be simulated and validated. Finally, a simulation model through Matlab will be also developed to compare and validate the results.
Prediction is basically about making claims of the future based on past information and current state. Predicting demand for the future can help many service organizations to adjust their resources thus reach their goals. In this paper, a complete framework for predicting workforce demand in service organizations using several techniques is provided. Moreover, two case scenarios of two service organizations requiring forecasting of demand are discussed. Also, this paper provides an initial test results of applying Moving average, Linear regression and Neural network techniques.
Accurate estimation of wind speed probability density distribution at a relevant wind generation site is crucial in maximizing the yield of the wind farm, and optimally utilize clean sources of energy. This goal calls for devising models with adaptable algorithms that accurately fit wind speed distributions regardless of the wind farm location and the distribution type. In this paper, the performance of the Kernel Density Estimation wind speed model is compared with that of a Gaussian Mixture Model, in which the optimal number of components of the Gaussian mixture model probability density function is obtained using the Bayesian Information Criterion approach.
Over the past few decades, interest in unmanned aerial vehicles and in particular quadcopters has increased due to the wide range of possible research applications that can benefit from the use of quadcopters. Insulator inspection on overhead lines has traditionally relied heavily on visual inspection. This task is both cumbersome and relies heavily on the experience of the inspector. It is also extremely dangerous as the inspector needs to work in close proximity with overhead lines, and contact with these lines can be fatal. This paper focuses on the development of a quadcopter based system that is able to inspect insulators on overhead power lines autonomously.
Envelope tracking is one of the methods used to enhance the efficiency of power amplifiers. Shaping function is a mathematical model used in envelop tracking that shows the relation between the biased voltage and the input power of the amplifier. In this paper, a new shaping function, that will allow the user to control the performance of the amplifier with the best possible efficiency, is introduced. The introduced shaping function could be used for any amplifier and could be modified to achieve any desired performance.
Currently, there is a large move towards 5G wireless technology beyond the existing, widely used 4G technology due to an increased use of smart devices, and multimedia content. 5G technology is expected to operate at high frequencies between 15 GHz and 100 Ghz opening up a new horizon for spectrum constrained future wireless communications. Designing high efficiency power amplifiers for such high frequencies presents a new challenge. This paper presents different designs of integrated PAs operating at 15-100 GHz for 5G applications.
This paper presents an area and power efficient multi-output switched capacitor (MOSC) DC-DC buck converter for energy-equality scalable SoCs including wearable biomedical devices. The MOSC converter has an input voltage range between 1.05V to 1.4V and generates two simultaneous regulated output voltages of 1V and 0.55V. The MOSC consists of two main blocks; a switched capacitor regulator and an adaptive time multiplexing (ATM) controller. The switched capacitor regulator generates a single regulated voltage using pulse frequency modulation based on a predetermined reference voltage. In addition, the ATM controller generates two simultaneous output voltages using pulse width modulation and eliminates the reverse current during the switching between the output voltages. Addressing the reverse current problem is important to reduce the voltage droop at the output resulting in a better performance.
In this paper we evaluate our proposed modification in the Stochastic Planning Using Decision Diagrams (SPUDD) to explicitly include and describe multi-objective problems. In order to test that we have used single objective Symbolic Perseus which is a Partially Observable Markov Decision Processes (POMDP) solver. We have created a parser that reads the multi-objective costs from the SPUDD and a scalarization function in Symbolic Perseus so it can be solved. In this paper we show promising results of experiments conducted to evaluate the parser and the scalarization function with different set of weights and compare it to the original SPUDD format.
Process mining is an emerging discipline that aims to analyze business processes using event data logged by IT system. Most of existing process mining techniques assume that there is a one-to-one mapping between process model activities and the events that are recorded during process execution. However, event logs and process model activities are at different level of granularity. In this paper, we present a machine learning-based approach to map low-level event logs to high-level activities. With this work, we can bridge the abstraction levels when labels are not available. The proposed approach consists of two main phases: automatic labeling and machine learningbased classification. In automatic labeling a modified k- prototypes clustering approach has been used in order to obtain the labeled examples. Then, in the second phase, we trained different machine learning classifiers using the obtained labeled examples. We verified our proposed approach using a real-world event log.
A method for reducing streaking artifacts in 4D-CT reconstruction by generating additional projections is proposed. The proposed method uses an Artificial Neural Network (ANN)-based interpolation algorithm for image generation. Deformable image registration algorithm is used to estimates the motion between the original images. Then, a multi-layer perceptron feedforward neural network with an adaptive learning procedure is used to interpolate the in-between images from original ones. Phantom and real-patient Computed Tomography (CT) scans are going to be used to test the algorithm. The generated images will be compared to the original ones to test the accuracy of the proposed algorithm.
